Printer-friendly versionPDF version

Li-ion batteries are now everywhere, in our phones, cameras, laptops, cars, watches and more recently hoverboards. You can also find them on much bigger systems such as ships and airplanes. Therefore, the demand for safe and efficient Li-ion batteries has never been as high as it is today: smartphone users want more autonomy without compromising on the device size, power tools users want to be able to get more power without compromising on the available energy, and fans of hoverboards want to ride them without fearing a fire … the list goes on.

Achieving the required safety and performance goals is very challenging for Li-ion battery designers because increasing the energy, power, specific energy or specific power of a cell is not trivial. It involves many parameters, such as the choice of the active material, current collectors and separators, each selected for their specific properties (e.g. thicknesses, porosities, electronic conductivities, etc.). What this means is that hundreds of these interdependent parameters need to be considered at the same time during the design process.

Large cell producers know this challenge very well because it costs a significant amount of money and time to test and validate all these different material combinations. Speeding up the design process and reducing cost is an objective all battery manufacturers have and Computer Aided Engineering (CAE) is part of the solution. It allows for testing multiple designs at very low cost in a fraction of the time usually required to perform experiments. In the race for cost reduction in cell manufacturing, CAE is the only way forward for future cell designs.

Battery Design Studio® (BDS), a powerful cell design software and cell testing platform, has a proven record that it is a reliable cell design virtual laboratory for Li-ion batteries. Indeed, with BDS, cell designers have access to all the tools they need to select and define in detail all the parts making up a battery cell. In addition, the accurate physics-based models offer users the ability to test a prototype under any operating condition and contribute to qualify the design.

In this way, hundreds of designs can be virtually tested at low cost. However, the building of each cell still takes some time and to achieve better designs, users rely on continuous analyses to help guide their design process. By coupling BDS to an intelligent design exploration tool such as HEEDS MDO®, the preparing of the virtual cells and running the analysis is automated. Furthermore, since HEEDS MDO enables design exploration, it intelligently drives the designs to reach the objectives defined by the engineer, such as increasing the energy density, or reducing the Ohmic resistance.

Next, I will demonstrate how a well-developed commercially available cell can still be further improved thanks to a unique coupling between BDS and HEEDS. In this case study, the focus is on increasing the energy density of the cell.

The Design Exploration Study

Before starting the optimization, it is important to accurately characterise the reference cell in BDS. It requires specification of the geometrical dimensions of each of the parts, such as electrodes, coating, tabs, etc, as well as the physics-based performance model to predict the cell behaviour. Only then can the design exploration/optimization be started.

Cell  specifications:

The battery cell in this case study is a common 18650 cylindrical shape, with a 1.5Ah capacity :

Figure 1: Cell specifications

The cathode active material is a blend of 20% Lithium Manganese Oxyde (LMO) and 80% Nickel Manganese Cobalt (NMC), which can be easily accounted for in BDS. It has a single tab design on both electrodes, with opposite side orientation. The cathode current collector is made of aluminium while the anode current collector is made of copper.

Figure 2: Cell being dissected to study tabs and electrodes designs

Tab designs, electrode dimensions and coating formulations can be easily input using the BDS friendly user interface. The details on the winding process can also be specified so that a well-controlled virtual cell design can be built.

Figure 3: BDS User Interface

Cell performance model:

The elaborate physics-based model is also defined and calibrated. The many controls enable accurate cell performance prediction (see picture below), comparing experimental and simulation voltage results on a 1C discharge procedure.


Figure 4: Cell voltage evolution during 1C discharge. Comparison between simulation (solid) and experiment (dots)


With this reference cell built, the optimization work can start. Since the objective is to maximise the energy density (Wh/kg), the changes will be focused towards weight reduction and increase in the coating length (in order to add more active material in the cell and therefore more energy).

The following design variables were selected for this study:

  • Positive and negative electrode tab positions

  • Positive and negative electrode current collector thicknesses

  • Tab width

  • Separator thickness

  • Separator porosity

Each of this design variables evolve within relevant ranges so they make sense from both physical and manufacturing standpoint.

Figure 5: Electrode dimensions parameters

Run and results analysis:

The optimizer drives the study to configure different designs in order to maximise the cell energy density (that is the available energy per unit mass, Whr/kg). After only 60 designs, the optimization reaches convergence, which means no significantly better designs could be found.

Figure 6: Monitor convergence - each dot represents a design. The highest the performance the highest the objective value.

Useful HEEDS MDO outputs allow the user to analyse trends and select the top best designs:

Figure 7: HEEDS lists all the designs, sorted by highest energy density here

The best design shows a ~9% increase of the energy density, that is an increase from 126 Whr/kg to 138 Whr/kg, which is a significant improvement on a cell design that was already well-optimized. Figure 8 shows the best design highlighted in black.

Figure 8: Best design in black - User can visually see the parameters variations across the study and observe trends

Figure 9: Best 4 Designs - Very close performance outputs, but quite different parameters combinations

Figure 9 shows the best 4 designs. It is interesting to note that although these designs are very close in terms of performance, they show quite a different combination of parameters. This allows the user to do a review to make a smart decision on the most appropriate combinations (for example : which of the four best performers has the lowest cost).

Additional design variables could be introduced, such as the tab designs (multiple tabs, edge collectors etc), electrode chemistries, electrolyte formulations and many more. Cell design is a highly complex problem with inter-dependent parameters, which BDS coupled to HEEDS can handle easily.

Additional applications of coupling BDS-HEEDS MDO

In addition to design exploration/optimization, coupling BDS to HEEDS MDO can also be used to perform curve fitting. This can be very useful in the case of calibrating a physics-based model, where sometimes some physical properties, like the solid diffusion coefficient, are missing because of their difficulty to be characterised experimentally. BDS and HEEDS MDO can then be used to fine tune the physics-based model by running different parameters combinations to fit the simulation outputs to cell test data.


This work was presented at the STAR Global Conference 2016 in Prague. For more information, please have a look at the presentation here.

I am thankful to my colleagues Dr. Robert Spotnitz and Dr. Taiki Matsumura who provided expertise that greatly assisted the completion of this work.





Matthew Godo
STAR-CCM+ Product Manager
Stephen Ferguson
Marketing Director
Brigid Blaschak
Communications Specialist
James Clement
STAR-CCM+ Product Manager
Dr Mesh
Meshing Guru
Joel Davison
Lead Product Manager, STAR-CCM+
Sabine Goodwin
Director, Product Marketing
Deborah Eppel
Technical Marketing Engineer